Document 3NyDxqe9NyoD3zezo8Y4zMegn

INVIROVMINTAl RtSIARt H 31. 374-380 (19831 & X * 'f . Vr- :* V V -* -^i-^ 'V^;r>. *'?>.."-'v'- ji>. *,* ?:'' > . "i .yrt-;c.. ;V-.- lu,, t-'i.-;-1 ^ &**>:? '-* - * ^aJ: , ^ ^ L*r-*k'A 3&i5i&e.:*t SWag5$* Xte:'**SZ %$>>?* Observations on the Distribution of Asbestos Fibers in Human Lungs1 THIS MATERIAL MAY BE PROTECTED BY CCPYRiGI-T LAW (TITLE 17, U. S. CODE) Andrew Churg5 and Patricia Wood Department of Pathology. Universirv of British Columbia. Vancouver. British Columbia. Canada Received July 39. 1983 In an attempt to begin evaluating the distribution of mineral fibers within lung, multiple small adjacent samples w ere selected from peripheral portions of three normal and six asbestolic lungs, and analyzed for asbestos and nonasbestos fiber content. For both normal and fibrotic lungs there was marked variation in fiber concentration from site to site: the mean in-case ratio of maximum to minimum fiber concentration was 7.3 foramosite and crocidolite. 7.0 for tremolite. 3.8 for chrysolite. 4.0 for nonasbestos fibers, and 3.4 for asbestos bodies. In order to test the precision oflhe membrane filter method of quantitating particulates, multi ple replicate samples of reference mineral standards were examined in a similar fashion: for these standard samples the ratio of maximum to minimum particle concentration varied from only 1.3 to 1.3. It is concluded that: (1) the membrane filler preparation technique used here for counting fibers provides reproducible results, and (3) there is marked variation in asbes tos and nonasbestos fiber concentration from area to area within lung tissue. The latter finding implies that anal>sis of small subpleural lung biopsies may produce misleading re sults. and that larger pieces or several pieces should be analyzed when attempting to define total asbestos fiber content of the lung. INTRODUCTION It is well established that asbestos and nonasbestos mineral particles are present in the lungs of every resident of industrialized countries, and that much larger numbers of asbestos fibers are present in those with occupational exposure to the dust (Churg. 1982). However, little is known about the distribution of any type of mineral particle within lung. Animal experiments have indicated that the initial site of asbestos fiber deposition is in the respiratory bronchioles and alveolar ducts (Brody et a!.. 1981) and that with time asbestos fibers tend to concentrate under the pleura (Morgan et al.. 1977). The few studies in humans which have looked at this problem have concluded that fibers are found in greater numbers near the visceral pleura (Sebastien et al.. 1977; Churg and Warnock, 1980); essen tially nothing is known about airway deposition. The differences described above occur over relatively large areas. Little is known about variations in fiber concentration within small areas. It is possible, for example, that the increase in concentration between central and subpleural sam ples of a given lung is more or less continuous, or equally possible that there are 1 Supported by a grant from the National Cancer Institute of Canada and Grants MA-6907 and MA-7830 from the Medical Research Council of Canada. X-Ray equipment used in this study was pur chased through a grant from the Worker's Compensation Board of British Columbia. 'To whom correspondence should be addressed: Department of Pathology. University of British Columbia. 3311 Wesbrook Mall. Vancouver. B.C.. Canada V6T IW5. 0013-9351/83 S3.00 Copyhfht I9S3 by Academic Press. Inc. All rights of reproduction in an) form resened 374 10003043 ASBESTOS FIBER DISTRIBUTION 375 distinct concentration jumps from one area to the next. Examination of histologic sections shows quite clearly that there are variations in the distribution of asbestos bodies, and we have previously suggested that in fibrotic lungs there are very marked concentration differences in numbers of bodies in different sites (Churg and Warnock, 1981). In this paper we begin to examine the question of mineral particle, especially asbestos fiber, distribution, by evaluating the concentrations of particles found in adjacent tissue samples. MATERIALS AND METHODS This study was based on lung tissue from nine individuals with a history of asbestos exposure, and from whom a long subpleural strip of lung tissue was available. Occupational and smoking histories, and histologic sections were avail able for all cases. Three cases were chosen in which the lung was grossly and microscopically normal; the remaining six cases were chosen because the lung was markedly fibrotic (asbestotic). For all cases a subpleura) strip of tissue measuring approximately 5 x 3 x 1cm was cut out. This was then divided into five 3 x 1 x 1-cm slices, and the slices further cut into three pieces of approximately 1 cc. One of the latter pieces was used for a preparation to count mineral particles by electron microscopy, one was used for a preparation to count asbestos bodies by light microscopy, and one was weighed, dried, and reweighed to determine a wet.dry weight ratio. Asbestos bodies and mineral fibers were extracted from lung tissue using meth ods we have previously described (Churg and Warnock, 1980). Briefly, this con sisted of dissolving the lung tissue in bleach, and, for the electron microscope preparations, further treating the sediment with hydrogen peroxide. The sediment was then collected on a Millipore filter of 0.45-Mm pore size. For light microscopy this filter was cleared and mounted on a glass microscope slide; for electron microscopy, pieces of the filter were cut out and the mineral particles transferred to coated electron microscope grids using the Jaffe washer technique. Asbestos bodies were counted by examining a known area of the Millipore filter under the light microscope and converting the result to numbers of bodies per gram dry lung. For determination of types and numbers of mineral fibers, 20 squares of each electron microscope grid were scanned at 18.000x and each fiber longer than 0.5 /im was examined for morphology, and by electron diffraction and energy dispersive x-ray spectroscopy to determine mineral identity. Results were converted to values per gram dry lung. For the purposes of this paper, amosite and crocidolite fibers were considered as one group. To determine the error in electron microscopic counting introduced by the method itself, a series of six replicate samples of U1CC amosite (as an example of an amphibole and a fiber) and six samples of cosmetic talc (as an example of a nonfibrous particulate) were created. These were prepared by suspending the reference mineral in water, taking six aliquots of known volume (the equivalent of taking six samples of lung), and processing them through bleach, peroxide, and Millipore filter steps as described above. Twenty squares were then counted for each sample, and the fibers or particles per milliliter calculated as above. 376 CHURG AND WOOD RESULTS Reference standards. Table I shows the counted values and the mean, standard deviation, range of counts, and ratio of highest to lowest count for the amosite and the talc standards. Excellent reproducibility was seen from sample to sample. The observed values for the amosite standard varied from 237- to 310 x 10" fibers ml. with a mean of 286 x 10s and standard deviation of 27 x 10. The ratio of highest to lowest observed value was 1.3. For the talc sample, values ranged from 159 to 238 with a mean and standard deviation of 198 and 37 x 10s particles/ml. The ratio of highest to lowest count was 1.5. Asbestos-exposed cases. Detailed results for the test cases are shown in Tables 2, 3, and 4. As a measure of the variation between sites w e have also calculated the ratio of highest to lowest observed value for each case, and have summarized these ratios in Table 5. For every type of fiber examined there was considerably more variation from site to site than w as present in the replicate samples of standard minerals. In a few: cases the observed ratio was nearly as low- as that for the reference standard (for example, nonasbestos fibers in case 1059 and case 1125. and amosite and crocidolite fibers in case 1012), but for the rest the ratios were much higher, reaching 16.5 for amosite and crocidolite in case 1054. The mean ratio for amosite and crocidolite over seven cases was 7.5. and for nonasbestos fibers over nine cases. 4.0. Similar variations were seen for chrysotile and tremolite (Table 5). For seven cases, a nonparametric correlation test (Spearman's p) was used to determine if the concentration of amphibole asbestos fibers was related to the concentration of nonasbestos fibers: the result was not significant (p = 0.21. P less than 0.56). Considerable variation was also found between highest and lowest ratio values of asbestos bodies (Table 4): these ranged from 1.5 to 15, with a mean ratio over seven cases of 5.4. In two cases asbestos bodies were not evaluated because of technical problems in sample preparation. A correlation test for asbestos bodies and fibers of amosite and crocidolite gave a nonsignificant result (p = -0.60. P less than 0.48). Sample 1 2 3 4 5 6 Mean SD Range Ratio" TABLE 1 Replicate Analyses of Reference Standards Amosite standard Talc standard 386 163 375 139 399 338 308 333 337 173 310 333 386 37 337-310 1.3 198 37 139-238 1.5 Sole. All values x 10* fibervml. * Ratio of highest lo lowest observed count. 10003049 378 CHURG AND WOOD S| c o Z ac 1 9*' V > ac c ' . t v<,* xv 1) - |c zo <* SC **! OS rs. ac C tr' v\ > vv V - e |- -c5 ~ 6 '5 <? < ac * -- ". -- *ri ^ I f**. U .? -e i2-?i = < N r. N ^ N tr, rl - V t i > ! s? -- c* h/. v% l/*Nr<C;eac 1 S ^ -- ce X s* -- t'i ^ Z 3? Vi 1 2 -J v 5 | s <E *25 i = il C h t * K r, t OC W ^ V"'. ^ OC ! fs ^ -- r -- Coc^^*nn<r t----W---t--- t N- T1 < ^ 4 *. * e c *2 "J 8 * r*. v. ^ ^ ri 3C J ri c I JQ jjj *wa -!i C Z J~| vnl c K f*> ^Cb <V 5 * * * >* * *% r at e : * <r. f*l N n n vC r. . *\ ^ *". ; *r. ^ S C !2 ; r. W ri t rl < n < r S . im *N r. ** 1 22 >. -5 S 96 & c* i n5 >x V 2v I CE JL*J5 , _ *T cS>* < 2 *, Z a ac ac pt i sj *, 2 s/s ft. fiC If 10003046 .1 ASBESTOS FIBER DISTRIBUTION TABLE 4 Asbestos Bodies Normal lungs Fibrotic lungs Case: Sample 1059 1012 1117 1066 1070 1025 1 2 3 4 5 Mean SD Range Rali<y 4 34 4 6 11 12 13 4-34 8.5 44 45 40 32 3 33 17 3-45 15 7.2 5.0 3.6 5.5 2.4 4.7 1.8 2.4-7.2 3.0 320 380 ND 390 260 340 60 260-390 1.5 31 45 48 37 12 35 14 12-48 4.0 1700 930 1000 700 1000 1000 380 700-1700 2.4 Sore. All values x 10s bodies/g dry lung. \ Ratio of highest to lowest observed value. 379 1125 440 660 200 410 320 410 170 200 - 660 3.3 DISCUSSION In the Introduction we have summarized much of the available data on variation of asbestos and other mineral fiber and particle distribution within the lung: it is clear that little information is available on this topic. Questions about asbestos fiber distribution are of particular interest, since asbestos-induced disease tends to have rather distinct localizations: asbestosis is usually w'orse in the lower subpleural zones. Asbestos shows a distinct tendency to produce disease in the pleura: this may appear in the form of effusions, fibrosis, plaques, or mesothelioma (Selikoff and Lee, 1978). Asbestos also causes peribronchial fibro sis (Craighead et a!., 1982) as well as bronchogenic carcinoma, lesions probably related to airway deposition of fibers. We have also recently shown that persons exposed to inorganic nonasbestos dusts may develop disease of the small airways, again implying the importance of particle localization (Churg and Wright, 1982). In this paper we have examined a method used by us and others for quantitating fibers and particles, and have considered the variation in particle distribution over small areas of lung. The data in Table 1 indicate that concentration variations TABLE 5 Mean Values of the in-Case Ratios of Highest to Lowest Observed Count Number of cases Mean ratio Asbestos bodies Amosite/crocidolite Tremolite Chrysolite Nonasbestos Fibers Standard amosite Standard talc 7 7 2 9 6* 6* 5.4 7.5 7.0 3.8 4.0 1.3 1.5 1 Number of samples. mm. rr 'J- ,< M&' 10003047 j. *,, i - . 380 CHUSG AND WOOD introduced by the dissolution, centrifugation, and membrane filter collection technique are satisfactorily small, both for fibers and for nonfibrous particulates. The standard deviations for the asbestos and talc standards are approximately 10 to 20C2 of the mean values, and the ratio of highest to lowest observed count is 1.3 to 1.5. Values for this ratio of much greater than 1.5 can be presumed, therefore, to reflect variations in the actual concentration in lung samples. It should be noted, however, that these conclusions apply to particulates which do not have a tendency to aggregate tightly: aggregated particulates are likely to produce much greater errors arising from uneven distribution on the electron microscope grid. As shown in Tables 2-4, most lungs demonstrate marked concentration differ ences from area to area for both asbestos and nonasbestos fibers, although a small number of lungs demonstrate fairly even distributions. This is also true of the concentration of asbestos bodies. There is a suggestion from the data that there is more variation in fibrotic lungs, but the numbers of cases are too few to be statistically significant. The results of the correlation tests indicate that the differ ent fiber types appear to vary independently in concentration: this is also true of asbestos bodies and fibers. These findings suggest that when studies are performed using digestion and membrane filtration techniques, as large a piece of lung as possible should be analyzed to minimize local site variation. They also suggest that small peripheral biopsies, i.e.. the kind of biopsy w hich is exactly equivalent in size and location to the samples we have examined, may be somewhat misleading if used to determine particle concentration. Certainly it would seem preferable to analyze at least two different pieces to minimize this type of error, as can be seen by selecting any 2 values for cases in Tables 2 and 3. REFERENCES Brody. A. R-. Hill. L. H.. Adkins. B.. and O'Connor. R. W. (1981). Chrysotile asbestos inhalation in rats: Deposition pattern and reaction of alveolar epithelium and pulmonary macrophages. Amer. /lev. Respir. Dis. 123. 670- 679. Churg. A. (1982). Fiber counting and analysis in the diagnosis of asbestos-related disease. Human Pathol 14, 381-392. Churg. A., and Wamock. M. L. (1980). Asbestos fibres in the general population. Amer. Rev. Respir. Dis. 122, 669- 678. Churg. A., and Wamock. M. L. (1981). Asbes(os and other ferruginous bodies: Their formation and clinical significance. Amer. J. Pathol. 102, 447- 456. Churg. A., and W right. J. L. (1982). Small airways disease in persons with exposure to nonasbestos mineral dusts. Hum. Pathol., in press. Craighead. 3.. Abraham. 3.. Churg. A.. Green. F.. Kleinerman. 3.. Seemayer. T.. Vallayaihan. N.. and Weill. H. (1982). Pathology standards for the diagnosis of asbestos-related diseases. Arch. Pathol., in press. Morgan. A.. Evans. 3. C.. and Holmes. A. 0977). Deposition and clearance of inhaled fibrous minerals in the rat. Studies using radioactive tracer techniques. In "Inhaled Particles IV. Part 2" i W. H. Walton and B. McGovern. Eds.), pp. 239- 274. Pergamon. New York. Sebastien. P.. Fondimare. A.. Bignon. 3.. Monchaux. G.. Desbordes. 3.. and Bonnaud. G. (1977). Topographic distribution of asbestos fibres in human lung in relation to occupational and nonoccupational exposure. In "Inhaled Panicles IV. Pan 2" (W. H. Walton and B. McGovern. Eds.i. pp. 433-444. Pergamon. New York. SelikofT. I. 3.. and Lee. D. H. K. (1978). "Asbestos and Disease." Academic Press. New York. i00030A8